On 10/13/2011 1:28 AM, Zsolt Kúti wrote:
Thu, 13 Oct 2011 03:47:09 +0200 -n
Иван Бишевац<[email protected]> írta:
Hi, I studied JavaSpaces technology and now want to write a
scientific paper that shows parallel computation with JavaSpaces,
using some algorithm. In book "JavaSpaces Principles Patterns and
Practice" I saw Manderbolt image comutation, but I want to show this
problem on some other example. Does someone have idea what problem to
take.
Hello Ivan,
I think good candidate can be any computationally expensive problem that
can be divided into parallely executable parts. The less overhead of
parallelization/networking has compared to the computational task , the
more economic and powerful the solution is. POV-ray or picture rendering
flashes to mind as scenic options to present nice presentations to this
problem space.
The scale of the computations involved has to dwarf the network transit times
before it gets interesting. Something on the remote machine has to provide an
automated scaling factor that provides a big benefit that a single machine can not.
For example, enterprise grade multi-core and multi-processor machines, can make
it convenient/possible to do some pretty large tasks on one machine in one
memory space. But, the cost of such machine can get quite high, quite quickly,
compared to the mid range desktop computer that you might use as a JavaSpace
client/worker computer. The network/JavaSpace mandated latency needs to be
considered so that you can partition the "job" into the right sized tasks so
that more "work" than "communications" is occurring.
Gregg
